Microsoft

Model · Fill mask

mdeberta-v3-base

Microsoft

DeBERTa improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. With those two improvements, DeBERTa out perform RoBERTa on a majority of NLU tasks with 80GB training data. In DeBERTa V3, we further improved the efficiency of DeBERTa using ELECTRA-Style pre-training with Gradient Disentangled Embedding Sharing. Compared to DeBERTa, our V3 version significantly improves the model performance on downstream tasks. You can find more technique details about the new model from our paper. Please check the official repository for more implementation details and updates. mDeBERTa is multilingual version of DeBERTa which use the same structure as DeBERTa and was…

Model · Image and text to text

Florence-2-base

Microsoft

This Hub repository contains a HuggingFace's transformers implementation of Florence-2 model from Microsoft. Florence-2 is an advanced vision foundation model that uses a prompt-based approach to handle a wide range of vision and vision-language tasks. Florence-2 can interpret simple text prompts to perform tasks like captioning, object detection, and segmentation. It leverages our FLD-5B dataset, containing 5.4 billion annotations across 126 million images, to master multi-task learning. The model's sequence-to-sequence architecture enables it to excel in both zero-shot and fine-tuned settings, proving to be a competitive vision foundation model. Use the code below to get started with the…

Model · Fill mask

deberta-v3-base

Microsoft

DeBERTa improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. With those two improvements, DeBERTa out perform RoBERTa on a majority of NLU tasks with 80GB training data. In DeBERTa V3, we further improved the efficiency of DeBERTa using ELECTRA-Style pre-training with Gradient Disentangled Embedding Sharing. Compared to DeBERTa, our V3 version significantly improves the model performance on downstream tasks. You can find more technique details about the new model from our paper. Please check the official repository for more implementation details and updates. The DeBERTa V3 base model comes with 12 layers and a hidden size of 768. It has only 86M…

Model · Image to 3d

TRELLIS-image-large

Microsoft

The image conditioned version of TRELLIS, a large 3D genetive model. It was introduced in the paper Structured 3D Latents for Scalable and Versatile 3D Generation.

Model · Image to 3d

TRELLIS.2-4B

Microsoft

TRELLIS.2 is a state-of-the-art large 3D generative model designed for high-fidelity image-to-3D generation. It leverages a novel "field-free" sparse voxel structure termed O-Voxel and a large-scale flow-matching transformer (4 Billion parameters). Unlike previous methods that rely on iso-surface fields (e.g., SDF, Flexicubes) which struggle with open surfaces or non-manifold geometry, TRELLIS can reconstruct and generate arbitrary 3D assets with complex topologies, sharp features, and full Physical-Based Rendering (PBR) materials—including transparency/translucency. - The CUDA Toolkit is needed to compile certain packages. Recommended version is 12.4. - Conda is recommended for managing…

Model · Fill mask

BiomedNLP-BiomedBERT-base-uncased-abstract

Microsoft

This model was previously named "PubMedBERT (abstracts)". You can either adopt the new model name "microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract" or update your transformers library to version 4.22+ if you need to refer to the old name. Pretraining large neural language models, such as BERT, has led to impressive gains on many natural language processing (NLP) tasks. However, most pretraining efforts focus on general domain corpora, such as newswire and Web. A prevailing assumption is that even domain-specific pretraining can benefit by starting from general-domain language models. Recent work shows that for domains with abundant unlabeled text, such as biomedicine, pretraining…

Model · Feature extraction

wavlm-large

Microsoft

The large model pretrained on 16kHz sampled speech audio. When using the model, make sure that your speech input is also sampled at 16kHz. Note: This model does not have a tokenizer as it was pretrained on audio alone. In order to use this model speech recognition, a tokenizer should be created and the model should be fine-tuned on labeled text data. Check out this blog for more in-detail explanation of how to fine-tune the model. - 60,000 hours of Libri-Light - 10,000 hours of GigaSpeech - 24,000 hours of VoxPopuli Authors: Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin…

Model

layoutlmv3-base

Microsoft

LayoutLMv3 is a pre-trained multimodal Transformer for Document AI with unified text and image masking. The simple unified architecture and training objectives make LayoutLMv3 a general-purpose pre-trained model. For example, LayoutLMv3 can be fine-tuned for both text-centric tasks, including form understanding, receipt understanding, and document visual question answering, and image-centric tasks such as document image classification and document layout analysis. Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, Furu Wei, ACM Multimedia 2022. If you find LayoutLM useful in your research, please cite the following paper: The content of this project itself is licensed under the…

Model · Object detection

table-transformer-structure-recognition

Microsoft

Table Transformer (DETR) model trained on PubTables1M. It was introduced in the paper PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents by Smock et al. and first released in this repository. Disclaimer: The team releasing Table Transformer did not write a model card for this model so this model card has been written by the Hugging Face team. The Table Transformer is equivalent to DETR, a Transformer-based object detection model. Note that the authors decided to use the "normalize before" setting of DETR, which means that layernorm is applied before self- and cross-attention. You can use the raw model for detecting the structure (like rows, columns) in tables.…

Model · Fill mask

deberta-v3-large

Microsoft

DeBERTa improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. With those two improvements, DeBERTa out perform RoBERTa on a majority of NLU tasks with 80GB training data. In DeBERTa V3, we further improved the efficiency of DeBERTa using ELECTRA-Style pre-training with Gradient Disentangled Embedding Sharing. Compared to DeBERTa, our V3 version significantly improves the model performance on downstream tasks. You can find more technique details about the new model from our paper. Please check the official repository for more implementation details and updates. The DeBERTa V3 large model comes with 24 layers and a hidden size of 1024. It has 304M…

Model · Feature extraction

wavlm-base-plus

Microsoft

The base model pretrained on 16kHz sampled speech audio. When using the model, make sure that your speech input is also sampled at 16kHz. Note: This model does not have a tokenizer as it was pretrained on audio alone. In order to use this model speech recognition, a tokenizer should be created and the model should be fine-tuned on labeled text data. Check out this blog for more in-detail explanation of how to fine-tune the model. - 60,000 hours of Libri-Light - 10,000 hours of GigaSpeech - 24,000 hours of VoxPopuli Authors: Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin…

Model · Table question answering

tapex-base-finetuned-wikisql

Microsoft

TAPEX was proposed in TAPEX: Table Pre-training via Learning a Neural SQL Executor by Qian Liu, Bei Chen, Jiaqi Guo, Morteza Ziyadi, Zeqi Lin, Weizhu Chen, Jian-Guang Lou. The original repo can be found here. TAPEX (Table Pre-training via Execution) is a conceptually simple and empirically powerful pre-training approach to empower existing models with table reasoning skills. TAPEX realizes table pre-training by learning a neural SQL executor over a synthetic corpus, which is obtained by automatically synthesizing executable SQL queries. TAPEX is based on the BART architecture, the transformer encoder-encoder (seq2seq) model with a bidirectional (BERT-like) encoder and an autoregressive…

Model · Image classification

resnet-50

Microsoft

ResNet model pre-trained on ImageNet-1k at resolution 224x224. It was introduced in the paper Deep Residual Learning for Image Recognition by He et al. Disclaimer: The team releasing ResNet did not write a model card for this model so this model card has been written by the Hugging Face team. ResNet (Residual Network) is a convolutional neural network that democratized the concepts of residual learning and skip connections. This enables to train much deeper models. This is ResNet v1.5, which differs from the original model: in the bottleneck blocks which require downsampling, v1 has stride = 2 in the first 1x1 convolution, whereas v1.5 has stride = 2 in the 3x3 convolution. This difference…

Model · Speech recognition

VibeVoice-ASR

Microsoft

VibeVoice-ASR is a unified speech-to-text model designed to handle 60-minute long-form audio in a single pass, generating structured transcriptions containing Who (Speaker), When (Timestamps), and What (Content), with support for Customized Hotwords and over 50 languages. - 60-minute Single-Pass Processing: Unlike conventional ASR models that slice audio into short chunks (often losing global context), VibeVoice ASR accepts up to 60 minutes of continuous audio input within 64K token length. This ensures consistent speaker tracking and semantic coherence across the entire hour. Users can provide customized hotwords (e.g., specific names, technical terms, or background info) to guide the…

Model · Fill mask

deberta-v3-small

Microsoft

DeBERTa improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. With those two improvements, DeBERTa out perform RoBERTa on a majority of NLU tasks with 80GB training data. In DeBERTa V3, we further improved the efficiency of DeBERTa using ELECTRA-Style pre-training with Gradient Disentangled Embedding Sharing. Compared to DeBERTa, our V3 version significantly improves the model performance on downstream tasks. You can find more technique details about the new model from our paper. Please check the official repository for more implementation details and updates. The DeBERTa V3 small model comes with 6 layers and a hidden size of 768. It has 44M backbone…

Model · Object detection

table-transformer-detection

Microsoft

Table Transformer (DETR) model trained on PubTables1M. It was introduced in the paper PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents by Smock et al. and first released in this repository. Disclaimer: The team releasing Table Transformer did not write a model card for this model so this model card has been written by the Hugging Face team. The Table Transformer is equivalent to DETR, a Transformer-based object detection model. Note that the authors decided to use the "normalize before" setting of DETR, which means that layernorm is applied before self- and cross-attention. You can use the raw model for detecting tables in documents. See the documentation for…

Model · Text to speech

VibeVoice-1.5B

Microsoft

VibeVoice is a novel framework designed for generating expressive, long-form, multi-speaker conversational audio, such as podcasts, from text. It addresses significant challenges in traditional Text-to-Speech (TTS) systems, particularly in scalability, speaker consistency, and natural turn-taking. A core innovation of VibeVoice is its use of continuous speech tokenizers (Acoustic and Semantic) operating at an ultra-low frame rate of 7.5 Hz. These tokenizers efficiently preserve audio fidelity while significantly boosting computational efficiency for processing long sequences. VibeVoice employs a next-token diffusion framework, leveraging a Large Language Model (LLM) to understand textual…

Model · Image classification

swinv2-tiny-patch4-window16-256

Microsoft

Swin Transformer v2 model pre-trained on ImageNet-1k at resolution 256x256. It was introduced in the paper Swin Transformer V2: Scaling Up Capacity and Resolution by Liu et al. and first released in this repository. Disclaimer: The team releasing Swin Transformer v2 did not write a model card for this model so this model card has been written by the Hugging Face team. The Swin Transformer is a type of Vision Transformer. It builds hierarchical feature maps by merging image patches (shown in gray) in deeper layers and has linear computation complexity to input image size due to computation of self-attention only within each local window (shown in red). It can thus serve as a general-purpose…

Model · Text to speech

VibeVoice-Realtime-0.5B

Microsoft

VibeVoice-Realtime is a lightweight real‑time text-to-speech model supporting streaming text input and robust long-form speech generation. It can be used to build realtime TTS services, narrate live data streams, and let different LLMs start speaking from their very first tokens (plug in your preferred model) long before a full answer is generated. It produces initial audible speech in ~300 ms (hardware dependent). ▶ Watch demo video (Launch your own realtime demo via the websocket example in Usage) Although the model is primarily built for English, we found that it still exhibits a certain level of multilingual capability—and even performs reasonably well in some languages. We provide nine…

Model · Image to text

trocr-small-handwritten

Microsoft

TrOCR model fine-tuned on the IAM dataset. It was introduced in the paper TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models by Li et al. and first released in this repository. The TrOCR model is an encoder-decoder model, consisting of an image Transformer as encoder, and a text Transformer as decoder. The image encoder was initialized from the weights of DeiT, while the text decoder was initialized from the weights of UniLM. Images are presented to the model as a sequence of fixed-size patches (resolution 16x16), which are linearly embedded. One also adds absolute position embeddings before feeding the sequence to the layers of the Transformer encoder. Next, the…

Model · Object detection

table-transformer-structure-recognition-v1.1-all

Microsoft

Table Transformer (TATR) model trained on PubTables1M and FinTabNet.c. It was introduced in the paper Aligning benchmark datasets for table structure recognition by Smock et al. and first released in this repository. Disclaimer: The team releasing Table Transformer did not write a model card for this model so this model card has been written by the Hugging Face team. The Table Transformer is equivalent to DETR, a Transformer-based object detection model. Note that the authors decided to use the "normalize before" setting of DETR, which means that layernorm is applied before self- and cross-attention. You can use the raw model for detecting tables in documents. See the documentation for more…

Model · Token classification

llmlingua-2-bert-base-multilingual-cased-meetingbank

Microsoft

This model was introduced in the paper LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression (Pan et al, 2024). It is a BERT multilingual base model (cased) finetuned to perform token classification for task agnostic prompt compression. The probability $p{preserve}$ of each token $xi$ is used as the metric for compression. This model is trained on the extractive text compression dataset constructed with the methodology proposed in the LLMLingua-2, using training examples from MeetingBank (Hu et al, 2023) as the seed data. You can evaluate the model on downstream tasks such as question answering (QA) and summarization over compressed meeting transcripts…

Model · Fill mask

deberta-base

Microsoft

DeBERTa improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. It outperforms BERT and RoBERTa on majority of NLU tasks with 80GB training data. Please check the official repository for more details and updates. We present the dev results on SQuAD 1.1/2.0 and MNLI tasks. If you find DeBERTa useful for your work, please cite the following paper

Model · Image to text

trocr-base-printed

Microsoft

TrOCR model fine-tuned on the SROIE dataset. It was introduced in the paper TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models by Li et al. and first released in this repository. Disclaimer: The team releasing TrOCR did not write a model card for this model so this model card has been written by the Hugging Face team. The TrOCR model is an encoder-decoder model, consisting of an image Transformer as encoder, and a text Transformer as decoder. The image encoder was initialized from the weights of BEiT, while the text decoder was initialized from the weights of RoBERTa. Images are presented to the model as a sequence of fixed-size patches (resolution 16x16), which…

Model · Fill mask

BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext

Microsoft

This model was previously named "PubMedBERT (abstracts + full text)". You can either adopt the new model name "microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext" or update your transformers library to version 4.22+ if you need to refer to the old name. Pretraining large neural language models, such as BERT, has led to impressive gains on many natural language processing (NLP) tasks. However, most pretraining efforts focus on general domain corpora, such as newswire and Web. A prevailing assumption is that even domain-specific pretraining can benefit by starting from general-domain language models. Recent work shows that for domains with abundant unlabeled text, such as…

Model · Token classification

llmlingua-2-xlm-roberta-large-meetingbank

Microsoft

This model was introduced in the paper LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression (Pan et al, 2024). It is a XLM-RoBERTa (large-sized model) finetuned to perform token classification for task agnostic prompt compression. The probability $p{preserve}$ of each token $xi$ is used as the metric for compression. This model is trained on the extractive text compression dataset constructed with the methodology proposed in the LLMLingua-2, using training examples from MeetingBank (Hu et al, 2023) as the seed data. You can evaluate the model on downstream tasks such as question answering (QA) and summarization over compressed meeting transcripts using…

Model · Image to text

trocr-base-handwritten

Microsoft

TrOCR model fine-tuned on the IAM dataset. It was introduced in the paper TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models by Li et al. and first released in this repository. Disclaimer: The team releasing TrOCR did not write a model card for this model so this model card has been written by the Hugging Face team. The TrOCR model is an encoder-decoder model, consisting of an image Transformer as encoder, and a text Transformer as decoder. The image encoder was initialized from the weights of BEiT, while the text decoder was initialized from the weights of RoBERTa. Images are presented to the model as a sequence of fixed-size patches (resolution 16x16), which…

Model · Image to text

kosmos-2-patch14-224

Microsoft

This Hub repository contains a HuggingFace's transformers implementation of the original Kosmos-2 model from Microsoft. Use the code below to get started with the model. This model is capable of performing different tasks through changing the prompts. First, let's define a function to run a prompt. Here are the tasks Kosmos-2 could perform: Once you have the entities, you can use the following helper function to draw their bounding bboxes on the image

Model

prophetnet-large-uncased

Microsoft

Pretrained weights for ProphetNet. ProphetNet is a new pre-trained language model for sequence-to-sequence learning with a novel self-supervised objective called future n-gram prediction. ProphetNet is able to predict more future tokens with a n-stream decoder. The original implementation is Fairseq version at github repo. This pre-trained model can be fine-tuned on sequence-to-sequence tasks. The model could e.g. be trained on headline generation as follows

Model · Video classification

xclip-base-patch32

Microsoft

X-CLIP model (base-sized, patch resolution of 32) trained fully-supervised on Kinetics-400. It was introduced in the paper Expanding Language-Image Pretrained Models for General Video Recognition by Ni et al. and first released in this repository. This model was trained using 8 frames per video, at a resolution of 224x224. Disclaimer: The team releasing X-CLIP did not write a model card for this model so this model card has been written by the Hugging Face team. X-CLIP is a minimal extension of CLIP for general video-language understanding. The model is trained in a contrastive way on (video, text) pairs. This allows the model to be used for tasks like zero-shot, few-shot or fully…

Model · Video classification

xclip-base-patch16-zero-shot

Microsoft

X-CLIP model (base-sized, patch resolution of 16) trained on Kinetics-400. It was introduced in the paper Expanding Language-Image Pretrained Models for General Video Recognition by Ni et al. and first released in this repository. This model was trained using 32 frames per video, at a resolution of 224x224. Disclaimer: The team releasing X-CLIP did not write a model card for this model so this model card has been written by the Hugging Face team. X-CLIP is a minimal extension of CLIP for general video-language understanding. The model is trained in a contrastive way on (video, text) pairs. This allows the model to be used for tasks like zero-shot, few-shot or fully supervised video…

Model · Image to text

trocr-large-printed

Microsoft

TrOCR model fine-tuned on the SROIE dataset. It was introduced in the paper TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models by Li et al. and first released in this repository. Disclaimer: The team releasing TrOCR did not write a model card for this model so this model card has been written by the Hugging Face team. The TrOCR model is an encoder-decoder model, consisting of an image Transformer as encoder, and a text Transformer as decoder. The image encoder was initialized from the weights of BEiT, while the text decoder was initialized from the weights of RoBERTa. Images are presented to the model as a sequence of fixed-size patches (resolution 16x16), which…

Model · Image to text

trocr-large-handwritten

Microsoft

TrOCR model fine-tuned on the IAM dataset. It was introduced in the paper TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models by Li et al. and first released in this repository. Disclaimer: The team releasing TrOCR did not write a model card for this model so this model card has been written by the Hugging Face team. The TrOCR model is an encoder-decoder model, consisting of an image Transformer as encoder, and a text Transformer as decoder. The image encoder was initialized from the weights of BEiT, while the text decoder was initialized from the weights of RoBERTa. Images are presented to the model as a sequence of fixed-size patches (resolution 16x16), which…

Model · Text to speech

speecht5_tts

Microsoft

SpeechT5 model fine-tuned for speech synthesis (text-to-speech) on LibriTTS. This model was introduced in SpeechT5: Unified-Modal Encoder-Decoder Pre-Training for Spoken Language Processing by Junyi Ao, Rui Wang, Long Zhou, Chengyi Wang, Shuo Ren, Yu Wu, Shujie Liu, Tom Ko, Qing Li, Yu Zhang, Zhihua Wei, Yao Qian, Jinyu Li, Furu Wei. SpeechT5 was first released in this repository, original weights. The license used is MIT. Motivated by the success of T5 (Text-To-Text Transfer Transformer) in pre-trained natural language processing models, we propose a unified-modal SpeechT5 framework that explores the encoder-decoder pre-training for self-supervised speech/text representation learning. The…

Model · Image segmentation

beit-large-finetuned-ade-640-640

Microsoft

BEiT model pre-trained in a self-supervised fashion on ImageNet-21k (14 million images, 21,841 classes) at resolution 224x224, and fine-tuned on ADE20k (an important benchmark for semantic segmentation of images) at resolution 640x640. It was introduced in the paper BEIT: BERT Pre-Training of Image Transformers by Hangbo Bao, Li Dong and Furu Wei and first released in this repository. Disclaimer: The team releasing BEiT did not write a model card for this model so this model card has been written by the Hugging Face team. The BEiT model is a Vision Transformer (ViT), which is a transformer encoder model (BERT-like). In contrast to the original ViT model, BEiT is pretrained on a large…

Model · Image to text

trocr-small-printed

Microsoft

TrOCR model fine-tuned on the SROIE dataset. It was introduced in the paper TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models by Li et al. and first released in this repository. The TrOCR model is an encoder-decoder model, consisting of an image Transformer as encoder, and a text Transformer as decoder. The image encoder was initialized from the weights of DeiT, while the text decoder was initialized from the weights of UniLM. Images are presented to the model as a sequence of fixed-size patches (resolution 16x16), which are linearly embedded. One also adds absolute position embeddings before feeding the sequence to the layers of the Transformer encoder. Next…

Model · Image to text

git-base

Microsoft

GIT (short for GenerativeImage2Text) model, base-sized version. It was introduced in the paper GIT: A Generative Image-to-text Transformer for Vision and Language by Wang et al. and first released in this repository. Disclaimer: The team releasing GIT did not write a model card for this model so this model card has been written by the Hugging Face team. GIT is a Transformer decoder conditioned on both CLIP image tokens and text tokens. The model is trained using "teacher forcing" on a lot of (image, text) pairs. The goal for the model is simply to predict the next text token, giving the image tokens and previous text tokens. The model has full access to (i.e. a bidirectional attention mask…

Model · Image segmentation

beit-base-finetuned-ade-640-640

Microsoft

BEiT model pre-trained in a self-supervised fashion on ImageNet-21k (14 million images, 21,841 classes) at resolution 224x224, and fine-tuned on ADE20k (an important benchmark for semantic segmentation of images) at resolution 640x640. It was introduced in the paper BEIT: BERT Pre-Training of Image Transformers by Hangbo Bao, Li Dong and Furu Wei and first released in this repository. Disclaimer: The team releasing BEiT did not write a model card for this model so this model card has been written by the Hugging Face team. The BEiT model is a Vision Transformer (ViT), which is a transformer encoder model (BERT-like). In contrast to the original ViT model, BEiT is pretrained on a large…

Model · Video classification

xclip-base-patch32-16-frames

Microsoft

X-CLIP model (base-sized, patch resolution of 32) trained fully-supervised on Kinetics-400. It was introduced in the paper Expanding Language-Image Pretrained Models for General Video Recognition by Ni et al. and first released in this repository. This model was trained using 16 frames per video, at a resolution of 224x224. Disclaimer: The team releasing X-CLIP did not write a model card for this model so this model card has been written by the Hugging Face team. X-CLIP is a minimal extension of CLIP for general video-language understanding. The model is trained in a contrastive way on (video, text) pairs. This allows the model to be used for tasks like zero-shot, few-shot or fully…

Model · Object detection

conditional-detr-resnet-50

Microsoft

Conditional DEtection TRansformer (DETR) model trained end-to-end on COCO 2017 object detection (118k annotated images). It was introduced in the paper Conditional DETR for Fast Training Convergence by Meng et al. and first released in this repository. The recently-developed DETR approach applies the transformer encoder and decoder architecture to object detection and achieves promising performance. In this paper, we handle the critical issue, slow training convergence, and present a conditional cross-attention mechanism for fast DETR training. Our approach is motivated by that the cross-attention in DETR relies highly on the content embeddings for localizing the four extremities and…